Please enable JavaScript for this site to run correctly.

Routable Active Travel Infrastructure Networks (RATIN)


To meet the UK's target to decarbonise transport and achieve net zero carbon emissions by 2050, tackle local air pollution, and address the obesity public health crisis and deliver well-being benefits, there needs to be a step-change in the use of active travel modes (Department for Transport, 2020). However, an overarching issue facing effective Active Travel use, innovation and planning is that there is no comprehensive, connected network or data framework available for active travel routing and personalisation. Whilst there has been a strong focus on urban design and infrastructure to support active travel there has not been the commensurate input to the network data. Spatial data is as much a part of the infrastructure needed to support active travel growth as the physical route and service provision. This gap in data provision poses a significant challenge to the successful delivery of active transport policies and planning and an opportunity to develop and populate an effective and standardised data model for GB.

active travel graphic

The flagship transport network dataset provided by Ordnance Survey, MasterMap Highways Network, consists of a road centre line with information on routing restrictions relevant to motorised modes. It can be augmented by the Paths product which contains footpaths and alleyways within larger urban areas (greater than 5 km2), but it primarily serves the interests of motorised transport. It is inadequate for pedestrian routing and analysis as it has no information on the presence (or not) of pavements. Consequently, route planning services that use the network can suggest unsuitable and dangerous routes for pedestrians. In addition, analysis of access to services and facilities can produce misleading results as it is implicitly assumed that every road (other than those where walking is expressly forbidden, such as motorways) is safe to walk on. These deficiencies are particularly unfortunate given the wide range of use cases for routable pedestrian networks within the public sector, as such networks are a crucial tool in effectively understanding and planning for active travel. Local authorities need a network product that can support them in identifying where the barriers to active travel are located, such as an absence of pavements, or insufficient pavement space and capacity to share with other modes (such as cycling and e-scooters). Accurate information on the suitability of roads for safe pedestrian use is vital if initiatives such as Mobility-as-a-Service are to effectively incorporate active travel modes either as standalone options or as part of combo travel (encouraging and enabling active travel as part of everyday journeys and commutes). Such information is also crucial if truly inclusive travel information is to be delivered which takes into account the requirements of those with differing mobility capabilities.

Phase 1

An overarching issue facing effective Active Travel use and planning is that there is no comprehensive, connected network or data framework for active travel routing and personalisation. The first phase of the RATIN project was therefore to:

  • Explore how such an active travel network (for walking and cycling) can be generated as far as possible from existing and open datasets
  • Test how other route datasets can be integrated to enhance the network and to richly attribute and maintain an enhanced framework for active travel analysis and use
  • Evaluate the gaps and barriers to data integration and maintenance (including data update, licensing and quality control)
  • Test the integration of diverse datasets to allow specification of the processing steps needed to update and maintain the Active Travel Network Data
  • Develop recommendations for an improved data structure and system to overcome gaps and barriers

Phase 2 of the project was centred around the creation of a routable network dataset, by using Ordnance Survey MasterMap (OSMM) as a starting point. Pavement width polygons were initially identified and then attributed with width information using a series of geo-processing steps. This is an improvement on the current National Geographic Database (NGD), which does not include distinct pavement features for either side of a road network.

Phase 3 of the project has culminated in the development of a geospatial framework and richly attributed network to support Active Travel (AT) use-cases. Additional network sources were added to the pavement network created in Phase 2, in successive stages in order to create a contiguous, routable network. In order of addition, these were:

  • OS Pathlink / Connecting links
  • Cleaned data from the crowd-sourced OS Maps application
  • OSMM Highways. A road feature that is possible to walk along, (e.g. non-motorway) and is the only option available as a link in the network
  • 'Other non-pedestrianised', possibly road links that connect the network. These could be a small section of dirt track or tarmac feature that is otherwise not classified or identified in any of the above stages.
  • Computed links between 'virtual' nodes (as crossing points between pavements only)
  • Links at dedicated traffic-light (pedestrian) crossings

The network was then attributed further to include additional contextual information that is relevant to specific AT use-cases, such as pedestrian / wheeled access and safe routes to schools. For Winchester, these attributes were:

  • Gradient, derived from LiDAR-derived Digital Terrain Model (DTM)
  • Presence of Hampshire County Council (HCC) Right of Way
  • Presence of street lighting
  • Roads where average speed is above the speed limit

For Romsey, additional attributes were added relating to surface type of the network, using a variety of image processing techniques with high-resolution aerial photography, including Meta's AI framework

  • Tarmac / sealed surface
  • Block paving
  • Gravel
  • Vegetation

The web maps for Phase 2 and Phase 3 contain functionality to calculate the shortest distance between an origin and destination. However when certain preferences are chosen by the user (e.g. minimum pavement width, gradient, signalled crossings where available) the shortest available route to accommodate these choices may be longer than if these options were not selected. The web applications contain a series of demonstration use-cases to highlight the feasibility of the data integration steps taken, for selected areas of Hampshire, UK.